JPH0781811B2 - Method for detecting wall thickness change of hollow body conducting electricity - Google Patents

Abstract

PCT No. PCT/CH87/00153 Sec. 371 Date Jul. 25, 1988 Sec. 102(e) Date Jul. 25, 1988 PCT Filed Nov. 18, 1987 PCT Pub. No. WO88/04028 PCT Pub. Date Jun. 2, 1988.The method is based on the exploitation of the "skin effect" occuring in any current-conducting material as a function of the current frequency going through said material. A transformer (5) is supplied from a generator (2) and an amplifier (4) to produce the current intended to supply the tube (T) and to measure the voltage at the terminals of a shunt (6). This voltage which is characteristical of the energization current is amplified by an amplifier (8) and brought to the reference input of a phase correlation amplifier (1). The resistive component of the signal on the tube is then measured by said amplifier (1) by just measuring the signal phase component.

Description

The present invention relates to a method for detecting a thickness change in the wall of a hollow body made of a material which is an electrical conductor.

The problem of detecting corrosion of a tubular body is not only the case of a buried tubular body where corrosion appears on the outer surface, but also the tubular body where a corrosive substance flows to cause corrosion on the inner surface. In this one or the other case, it is not possible to simply carry out a visual inspection of the condition of the tubular body.

This is why different methods have been proposed based on magnetic or electrical parameters which make it possible to obtain a characteristic signal indicative of the state of corrosion of the wall of the tubular body.

US Pat. No. 4,048,558 already uses the "skin effect"("1'effet de peau"), whose depth lies at the frequency of the current, to detect the presence of defects in the wall of the tubular body. It has been proposed.

This method consists of supplying different frequencies of current to the surface of the tubular body and measuring the frequency-dependent change in the impedance of this tubular body.

Each such measurement is compared to either that portion of the tubular body to be measured, another portion of the same tubular body maintained at the same temperature, or either of the reference tubular bodies. This treatment method is based on the specific resistance, the metal of the tubular body and the magnetic permeability ρ due to its temperature.
It is possible to eliminate the .mu. And .mu. Parameters so that the changes appearing during the comparison process will necessarily be due to structural differences in the walls of the two compared tubular parts.

This processing method requires that the sample and reference tubular bodies have identical ρ and μ values. It has been found that this is not possible in the case of a buried tubular body, and no one can reproduce the same conditions on the reference tubular body as in the buried state. Therefore, in this case the parameters ρ and μ are not possible, as well as possibly unknown drawbacks.

European Patent (EP) 0,175,257 also proposes to detect defects in hollow conductors by using the skin effect.

According to this method, the hollow conductor is supplied with an alternating current having a predetermined frequency in a given direction, generates a magnetic flux having a given direction, and a voltage drop contacts the hollow conductor. A hollow conductor which is picked up through two measuring points which are separated from each other by a distance given by the two measuring lines, the first conducting loop having a useful guiding surface as small as the measuring line. And a second conductive loop is formed parallel to the direction of the current and perpendicular to the direction of the magnetic flux, with the second loop of the induced voltage caused by the change in the magnetic flux as close as possible to the measuring point. A measurement is made, the voltage drop and the induced voltage are supplied to an electronic evaluation circuit for measuring the voltage drop and the partial resistance of the hollow conductor between the measuring points is estimated from the voltage drop and the induced voltage.

This method is not suitable for measuring defects on tubular body parts of lengths of several meters, and for that reason means such as auscultation of the outer surface of the conductor are required. , Access over the entire length of the tubular conductor. And it is only possible to measure on very short sections and the auscultation device is required to be moved along a hollow tubular conductor.

In a buried tubular body, electrical measurements are difficult and inaccurate due to the effect of soil covering the tubular body. This is because the whole device is inserted into a tubular body (hereinafter simply referred to as a pipe) for measuring and recording the measured results, and the pipe is moved by a fluid flow or by an independent propulsion means. It is the reason why it has been proposed to mount it on a moving support intended to be pulled by either one. This technique uses a measurement method such as ultrasonic echography.

The analysis of courants de Foucault and magnetic field is complicated, and the search for defects requires precise knowledge of the displacement parameters of this moving support.

The present invention proposes a method for detecting a thickness change in the wall of a pipe, which in the case of an embedded pipe, the recording element progressively advances in the pipe. Accordingly, it enables signals to be transmitted to the outer surface of the pipe in such a way as to be able to track the evolution of the measured parameters depending on the thickness, which results in disadvantages. It is possible to obtain accurate detection and knowledge about its size.

However, in this method it is only necessary to have a moving pick-up element if one only wants to know the exact location where the defect is present, and the existence of a mere defect and its size is fixed. Can be picked up by an element.

To this end, the invention comprises a method according to claim 1.

The essential benefit of the present invention lies in its sensitivity and the fact that the measured signal is independent of the environment in which the pipe is present.

The signal, which is the characteristic voltage for the 1'impdance transversale of the pipe wall measured by the two electrodes, is caused by the conductor to make a direct correlation between the advance of the electrode and the change of the signal. Is transmitted to the outside of the pipe in such a way as to allow

The accompanying drawings illustrate, by way of example and in the form of diagrams, different modes for carrying out the method which is the subject of the present invention.

FIG. 1 is a diagram schematically showing a longitudinal portion of a tubular body, in which characteristic parameters in the method of the present invention are entered.

2 and 3 are diagrams related to the diagram of FIG.

FIG. 4 is a phase diagram (diagramme de phase) of the picked up signal.

FIG. 5 shows a diagram relating to the processing of this signal.

FIG. 6 is a block diagram of the measuring device.

7 and 8 are diagrams relating to the operating modes of the method of the invention.

FIG. 9 is a perspective view showing another different application of the method according to the invention.

The method of the invention is based on the use of the "1'effect de peau" which appears in any material that is an electrical conductor, depending on the frequency of the current flowing through the material.

In the following description, it is considered that the tubular body T shown in FIG. 1 is usually supplied with an auxiliary current I on its outer surface.

If one considers FIGS. 1 to 3 first, it will be seen that both the current density j and the electric field E change in the same way, depending on the frequency of the current.

In fact, the general formula for the current density j depending on the depth X in the wall of the pipe is as follows: j = j _o e ^{−K / a} where “a” corresponds to the depth of the skin effect Is expressed as follows: a = ζ / πμF (ζ is the specific resistance, μ is the permeability and F is the frequency) j _o is the current density on the outer surface of the pipe. Such current density decreases rapidly from the outer surface to the inner surface of the pipe T mainly as x> a.

This change is shown in the table below, where e ^{−K / a}
Values are shown for x _i = 0, a, 2a, 4a etc.

x _i ^{Ki / a} e ^{−K (i + 1) / a} / e ^{−Ki / a} 0 1.0 a 0.368 0.368 2a 0.135 0.368 4a 0.018 0.135 8a 3.35.10 ^-7 0.0183 16a 1.125.10 ^-17 3.35.10 ^-4 The electric field to be changed similarly changes according to the following equation: E = jz _s z _s : surface impedance = ωζμ (ω: pulsation du caurant d'excit
ation) = 2πF) Suitable at the inner surface of the pipe T at the portion Q _B where the wall thickness is reduced, as shown in the diagram of FIG. 2 forming the graphical representation of the table above. The ratio between the electric field E ₁ measured at the selected frequency and the electric field in the normal thickness part Q _A varies strongly depending on the wall thickness.

This is a phenomenon that the present invention proposes to utilize to show the existence of a thinning of the pipe wall.

The signal measured on the inner surface of the pipe and the corresponding electric field on that surface is a result of the inductive component, which is essentially partly dependent on the resistivity component and the induced tensions parasites induites. is there.

Thus, only the resistivity component is a good measure of the skin effect. Due to such facts, it is necessary that the measuring device intended for operating this method must be designed in such a way that it rejects the part of the signal that is not in phase with the current.

The result is a phase correlation amplifier (unamplificateur corr
relation de phase) or frequency and phase response analyzer (un
analyseur rponse en frquence et en phase)
Obtained by either Depending on the format of the embodiment,
An apparatus using the phase correlation amplifier will be described below. With reference to FIGS. 4 and 5, the mode of operating an amplifier used as a dmondlateur synchrone operating according to the following method will be explained.

All signals S (FIG. 4) may be thought of as the vector sum of two vectors, S _O , one in phase with the reference and the other, S ₉₀ , formed at right angles to this reference.

If the demodulator is correlated in phase with the reference signal R _S , the modified signal S _R appearing at the output will be a vector function only in phase with S ₀ , as can be seen on the diagram of FIG. 5, where Can subsequently be followed by a reference signal R _S which allows the formation of a rectangular signal F made of steel to modify the measurement signal S to obtain a correction signal S _R.

The block diagram of FIG. 6 shows the phase correlation amplifier 1 type PAR128A, frequency generator (u
n gnrateur de frquence) 2 types Wavetek 13
4, frequency counter 3 type HP5300A, power amplifier 4
Type Bruel and Kjaer 2706, transformer ratio 35 and maximum secondary current
20A toroidal transformer (un trans form atseur to
rodal) 5 Maximum current 10A, 10mΩ shunt 6, Voltmeter 7
Type Solartron 7040, low power amplifier 8 type Levell TA
The present invention relates to a device including the switch box 605 and the switch box 9.

The frequency may vary between 3 and 1000 HZ with a maximum current of 10 A.

The function of this device is to supply the current from the generator 2 via the amplifier 4 to the transformer 5, thereby producing a current intended to be supplied to the pipe T, and to reduce the voltage at the terminals of the shunt 6. It is something to measure. This voltage, which is characteristic of the current supplied, is first amplified 10 times by the amplifier 8 in order to obtain a sufficient amplitude and is then transmitted to the reference input of the phase correlation amplifier 1.

The output of the amplifier 8 is connected to the measurement input of the phase correlation amplifier 1 through an intermediate switch box 9. This last connection mode can allow not only phase control but also current amplification measurement and control.

The resistivity component of the signal on the pipe is then measured by connecting to the measurement points at the differential inputs A and B of the phase correlation amplifier 1 and by measuring only the in-phase component of that signal.

Such a connection must be constructed in such a way that only the internal electric field (le champ interne) is measured, which makes it possible to use a coaxial cable which is insensitive to external magnetic flux to exit the tubular body. It means that it is necessary.

The measurement of the signal itself picked up on the pipe T after the reference signal has been supplied is carried out in three different ways.

One measuring point may be fixed on the inner surface of the pipe T and the other point may be displaced longitudinally on this same surface. This displacement may be repeated several times while changing the frequency of the current supplied for each movement of the movable part.

The part Q _{B of} reduced thickness is made and the pipe T
The other part shows a constant thickness Q _A. Pipe part T
In the case of, the distance L between two points that are separated from each other
A diagram of the voltage V depending on is shown in FIG. It can be seen that for inspection of pipes the drawbacks become particularly apparent in the 100 HZ range. This measurement makes it possible to determine the size and the longitudinal position of the defect once and simultaneously.

According to another method of operation of this measuring process, the two points for measuring the signal relating to the inner surface of the pipe T are kept at a fixed distance from each other and both move along this inner surface. Be punished.

This movement may be repeated for one and the same part of the pipe, varying the frequency of the supply current which also constitutes the reference signal. The existence of defects is immediately apparent in the form of peaks which change in a frequency-dependent manner in the diagram of FIG. This peak is a maximum between 100HZ and 200HZ, for example. As in the previous case, the peak value is a characteristic value for the size and location of defects along the pipe.

The inconvenience in these solutions arises from the need to displace at least one signal pick-up point inside the tubular body. In the case of a buried pipe, this requires that the movable pick-up point or pick-up points be placed on a transfer member which obviously has to be inserted and substantially pulled into the pipe during the measurement. .

Such members intended for displacement in a hollow tube already exist, but the solution is based on the principle of the skin effect on which the present invention relies, but the measurement of the signal is fixed. What you can do with help may also be interesting. In these situations it is possible to fix two measuring points separated by a given distance on the inner surface of the pipe, and when they are coaxial by means such that they do not pick up the external magnetic flux. By means of a combination of two conductors arranged in a mode, it is possible to connect the measuring device to an external device to which it is connected.

The electric field present on the inner surface between these two points is then measured in a frequency-dependent manner.

In this variant for picking up the signal,
The recorded curve, which is dependent on the sweep frequency (du balayagede frquennce), tracks the detection of defects in the thickness of the pipe, with the aid of a computer, which traces the unknown parameters, namely the resistivity ζ and the permeability μ. This is made possible by comparison with a series of curves obtained by making changes in such non-geometric parameters.

Even if the depth of corrosion is arbitrarily fixed, the length of the corroded part is unknown and thus will again constitute one parameter that needs to be changed.

A comparison of these curves calculated by the measured curves will make it possible to ascertain the existence of defects and, within certain limits, to infer their magnitude. Ah On the other hand, the defect or defects between the two measuring points cannot be determined in this case.

Such disadvantages are compensated by the fact that such signal pick-up methods do not require the mobility of the electrodes inside the pipe, which obviously constitutes a simplification in the mode of operation.

In the case of the new device, the pipe may be provided with internal electrodes for picking up the signal, which are connected to the measuring part, for example by a cable system.

For example, in existing equipment for gas pipes in cities,
It is possible to have access to the interior of the pipeline by different branch connections connecting to the dwelling.

These same branch connections can be used to insert movable electrodes inside the pipe according to the pickup mode described above.

As mentioned above, the above description deals with the case where a measurement inside the tubular body is performed. If it is desired to carry out a measurement on the outer surface of the tubular body, a bias must be applied inside the tubular body, whereby the skin effect itself appears on the inner surface.

However, in the case of frequency-based electric field measurements between two fixed points, and as described above, the curves obtained were tracked with the help of a computer and obtained by varying unknown non-geometric parameters. Performing a measurement on the same side of the pipe when comparing it to a series of curves generated thereby makes it possible to use the same method by which the current supply is performed.

The signal measured next will in fact be a signal of the change in electric field depending on the frequency of the current sent through the pipe.

Therefore, measurements on buried pipes are possible by simply contacting the outer surfaces of the walls of these pipes. The different processing methods using such comparison operations will be described in some detail below. All parameters for the pipe are assumed to be known, except for changes in these parameters that are the result of corrosion.
On the other hand, the electrical parameters are unknown because no one is aware of the temperature or the electrical properties of the material from which the tubular body is made and the effects of the environment.

The mathematical model used is the impedance of the pipe (1) at DC or very low frequencies (<1HZ), (2)
At a high frequency where the skin effect is global, ie the skin effect itself appears completely in the range of the wall thickness of the pipe, and (3) the intermediate frequency at which the skin effect occurs over the thickness of the pipe wall. , Based on three appropriate formulas for calculating. These three equations are shown below in the above order.

Where Lo represents the length of the pipe of interest, Sa represents the cross-sectional area with outer radius r and inner radius q,
ζ represents the specific resistance.

Where ζ is the specific resistance, μ is the permeability and F is the frequency.

The third equation gives the impedance per unit length Z derived from HBDuright's "Exact Calculation Method for Skin Effects in Insulated Tubes".

And Where R _DC is the resistance of the pipe to DC. j is an imaginary number Is.

The depth d in the conventional skin effect is Is related to m.

f is the frequency, μ is the magnetic permeability of the conductor, ζ is the resistivity of the conductor, r is the outer radius of the pipe, q is the inner radius of the pipe, and ber and bei are the Kelvin functions of the first type (la fonction).
de kelvin) and ker and kei respectively represent the true and imaginary parts of the second type of Kelvin function, and all four values are zero Is a number.

bar ', bei' and kei 'are derivatives for the corresponding Kelvin functions.

Thus Lo, r and q are known geometrical parameters, ζ and μ are unknown parameters in addition to the outer radius of the pipe in the corroded area.

Many different processing methods may be required to create and analyze the correlation between the measured longitudinal impedance based on different supply current frequencies and such mathematical models.

Depending on the form of the embodiment, this problem may be solved. 3
We mention three different processing methods. In the first embodiment, the pipe section under test is completely free of corrosion, including a complete application of the model.

Equation (1) is the ζ corresponding to the uncorroded pipe
It is possible to calculate the value of.

Next, the equation (3) is used to calculate the value of the magnetic permeability μi for each frequency fi so that Z ′ / R _DC > 1.

If the value of the continuous μi calculated starting from this ζ is constant, or if it follows the law corresponding to the permeability of the metal used to manufacture the pipe, then Initial guesses that the pipe was not corroded proved to be confirmed. In the opposite case, it can be concluded that the pipe is corroded between the two measuring points. The following is a summary of the configuration of one specific example of the method for detecting a change in wall thickness of a hollow tubular body according to the present invention:
(A) preparing a mathematical model for calculating the impedance of the tubular body using biasing currents in at least two kinds of predetermined frequency regions in which the skin effect by the current is substantially different, (b) Supplying a reference alternating current to one surface of the wall of the tubular body, (c) changing the frequency of the reference alternating current according to the predetermined frequency range, (d) one of the tubular bodies Measuring, between two electrodes arranged on the surface and along the longitudinal direction, spaced apart from each other, a signal indicative of the characteristic voltage associated with the field effect at each frequency of the reference alternating current, And a step of excluding a signal part which is a part of the measured signal and which is not in phase with the reference AC current, (e) a step of arbitrarily estimating a change in the thickness of the wall portion, (f) an unknown parameter is specified In order to do the first Solving a first equation of the mathematical model using a measured signal measured in response to a frequency in the frequency domain of Eq. And an estimate for the thickness change, (g) another unknown In order to specify the parameters, the second
Solving the second equation of the mathematical model using the other measurement signal measured in response to one frequency in the frequency domain and the parameter obtained in the step, (h) relating to the thickness Checking changes in the other parameters to determine if the above values are correct,
And (i) if the above values for the thickness are not correct, the other steps for thickness variation are selected and then steps (e)-(h) above are repeated and the obtained thickness is This process is repeated until the change value is proved to be correct, and is a method for detecting a change in wall thickness of a tubular body made of an electric conductor material.

According to a second embodiment, the pipe comprises a limited geometrical corroded area, for example where the defects extend in the form of a ring on the outer surface of the pipe, the width of which is unknown. Is assumed to exist and it is presumed that the corrosion affects half the wall thickness.

Applied to the mathematical model for this new situation. In this case, the mutual relationship and its judgment include two variants, one is a simplified variant with low accuracy and the other is more complex with more accurate variants.

In the first variant, it is shown that the value of R _AC , which is related to the overall skin effect to determine the product of ζ · μ, is practically independent of the presence of corrosion defects. Equation (2) is used.

With the aid of equation (3), along with the aid of the sweep loop, the skin effect is limited to the wall thickness corresponding to the thickness remaining in the hypothetical corrosion (50%) with an annular distribution. To find the value of μ (?) Corresponding to the impedance value measured at the selected limited frequency.

It appears to be a corroded area with the aid of a second sweep loop, which may include those mentioned above, and with the aid of equation (3), which is adapted to rely on hypothetical geometrical imperfections. The width of the area is determined. And its width is
This corresponds to what the measured longitudinal impedance value is to be found again.

A drawback of this method arises from the fact that the drawbacks for the ζ · μ product at relatively high frequencies have to take into account the value of μ which may already be influenced by the frequency. Using one single frequency for the loop on μ simply uses some of the available information. Finally, the fact that the investigation is unilaterally based on one type of corrosion defects does not take into account the fact that pipe corrosion may be of another type.

The modal variation on this correlation uses a series of mathematical models for many different types of corrosion initiation. Furthermore, this correlation is, on the one hand, the frequency from the time when the effect of the skin effect appears, avoiding considering the results of measurements corresponding to high frequencies in such a way as to reduce the effect of changes in frequency-dependent μ. The dependent impedance changes must be taken into account.

The variant illustrated by FIG. 9 illustrates the application of the process according to the invention for the detection of crack propagation for tensile test sample 10.

Areas 10a of the test sample that are split due to pulling action are folded over on either side of the test sample 10 so that an electrical connection is formed between the test sample 10 and the hollow case 11. Integrated into a hollow metal case 11 via a deformable insert such as a copper braid 12 acting on the case, which case is related to the tensile force applied to the test sample. Not not.

When the method of the invention is carried out, case 11 with test sample 10 is supplied with current via one of the internal or external surfaces and the electric field on the other surface is measured. This electric field has been shown to depend on the lateral impedance as cracks develop and propagate on the test sample, so the measured electric field is higher at higher frequencies and the deeper the cracks are. To increase.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Kornmann Michel Zehher, Switzerland 1212 Gran-Ransi, Shemande Palette, 31 (72) Inventor Gordietel France, F-38500 Boiron, Avenue Eduard Erio, 5 (56) Reference JP-A-61-80039 (JP, A) US Patent 4048558 (US, A)

Claims (4)

[Claims] 1. A step of supplying a reference alternating current to one surface of a wall of a tubular body, the electric field between two electrodes longitudinally separated from each other on the other surface of the wall. Measuring a signal indicating a change in characteristic voltage of the signal, excluding a signal part that is a part of the measured signal and is not in phase with the reference alternating current, the distance between the electrodes, and the electrode is arranged along the tubular body. And a step of changing at least one parameter selected from the longitudinal position of the two electrodes and the frequency of the energizing current. A method for detecting changes in body wall thickness. 2. The method for detecting a change in wall thickness of a tubular body according to claim 1, wherein the frequency of the reference alternating current is 200 Hz or less. 3. (a) Prepare a mathematical model for calculating the impedance of the tubular body by using the biasing currents in at least two kinds of predetermined frequency regions in which the skin effect by the current is substantially different. (B) supplying a reference alternating current to one surface of the wall of the tubular body, (c) changing the frequency of the reference alternating current according to the predetermined frequency region, (d) Characteristic voltage for field effect at an individual frequency of the reference alternating current, on one surface of the tubular body and between two electrodes spaced apart from each other along the longitudinal direction. Measuring a signal and excluding a signal portion that is a part of the measured signal and is not in phase with the reference AC current; (e) arbitrarily estimating a change in the thickness of the wall portion; (f) Identify unknown parameters For solving the first equation of the mathematical model using one measurement signal measured in response to one frequency in the first frequency region and the estimated value regarding the thickness change, g) In order to identify other unknown parameters, the second
Solving the second equation of the mathematical model using the other measurement signal measured in response to one frequency in the frequency domain and the parameter obtained in the step, (h) relating to the thickness Checking changes in the other parameters to determine if the above values are correct,
And (i) if the above value for the thickness is not correct, then another value for the thickness change is selected and then steps (e)-(h) above are repeated, and the obtained thickness is A method for detecting a thickness change of a wall of a tubular body made of an electric conductor material, characterized in that this step is repeated until the change value is proved to be correct. 4. A characteristic voltage change with respect to the electric field between two electrodes which are supplied with an alternating current on one surface of the wall of the tubular body and are longitudinally separated from each other on the other surface of the wall. To
It is measured by changing the frequency of the energizing current, and the resulting curve is used to estimate the length of the portion where the resistivity and permeability as well as the thickness change with respect to the estimated change in thickness. A method for detecting wall thickness variation of a tubular body made of an electrically conductive material characterized by comparison with a series of curves obtained by varying and calculating the height.